It doesn't exactly answer the question, but I -- someone who doesn't watch video and wants text, articles -- have watched this video twice over. It's worth it :)
I was a bit skeptical but this is actually really neat. Sometimes they nail it. The situations where both are bad are common, and seeing which LLM produced them lines up quite well with how good I think the same models are.
Is it possible to summarise? I can't watch a video currently.
I'm also aware Göbekli Tepe is a favourite of pseudoarcheologists and conspiracy theorists. Given who Finkel is, it will be nice to see a theory presented by someone who is actually a genuine expert! :)
> Did this small stone represent the very beginnings of writing, about 12,000 years ago? And did archaeologists really overlook such an important find all this time? (Spoiler: No, it doesn’t. And we didn’t.)
Hmm, but later in the post, discussing neolithic symbols from the wider region:
> Interestingly, among these depictions there are some which indeed appear to be symbolic substitutes for more complex images, like the bucranium (ox head) in place of a full aurochs, or arrow-like zigzag lines representing snakes, and large birds reduced to a few characteristic lines. These depictions and their “abbreviations” seem to adhere to a certain convention, kind of a standardisation even, suggesting a communication system that uses these stone objects as media to store important information and knowledge.
But totally not writing, yo.
It's unfair to attack archaeologists, who only busy diligently doing their job, increasing the store of public knowledge, and winning awards from each other. Except sometimes they deserve it, so I'm going to attack them by saying that they hate to be specific. It's a huge risk, you might be wrong, or ridiculed as a crank, or you might attract cranks. Better to couch everything in language that lets you avoid saying the thing.
On the other hand we have the 40,000 year old mammoth carving from Germany (see Sci Am article in a different comment) with decorative cross patterns on it, touted as "statistically complex" and conveying information, when they're just badly carved decoration. In that case they were simultaneously pussyfooting around about stating their case and going off on a flight of fancy, at the same time, in the style of "I'm not saying it's aliens".
>> Interestingly, among these depictions there are some which indeed appear to be symbolic substitutes for more complex images, like the bucranium (ox head) in place of a full aurochs, or arrow-like zigzag lines representing snakes, and large birds reduced to a few characteristic lines. These depictions and their “abbreviations” seem to adhere to a certain convention, kind of a standardisation even, suggesting a communication system that uses these stone objects as media to store important information and knowledge.
>But totally not writing, yo.
Children drawings have the same characteristics but are not writings
Yes, and the final sentence "But in my humble opinion we’re not seeing phonetic values assigned to specific symbols representing spoken language here yet" is fair enough.
The letter A evolved from an ox head, but that's mere coincidence, showing only an enduring interest in symbols of ox heads.
I like that you keep it wholly on-device - that makes me feel more comfortable using it. Is there a way to point it at existing samples of my writing to help train it?
And can it be different per app?
For example, I have a novel. I would never AI write a novel but autocompletion is fairly reasonable. My writing there is notably different to an email. If it was active in Scrivener, I'd want it to be using Scriven writing samples for consistency.
I have never used them (don't trust them to accurately capture what is important in a meeting vs just noting what's mentioned), but the concept seems very useful to me.
Reminds me of when I worked for a small shop which had the copier maintenance contract at a local college --- when something went wrong and wasn't properly addressed, my bosses found themselves being held to account with their own words from prior phone calls being quoted back to them verbatim --- which they were mystified by until I explained that the administrators had all come up from the clerical pool and knew shorthand.
The main risk is attorney client privilege and it's already been tested in New York, if you transcribe a call you need to turn over the transcriptions and they can subpoena the company doing the transcription for the records if you refuse.
>they want to keep attorney client privilege so they can outright lie
As a trial attorney for over 40 years, that is an incredibly offensive take. Attorney/client privilege is usually litigation related and a prime example of the nature of conversations involve our advising our client of the prospects of prevailing at trial and whether to engage in settlement discussions with money amounts involved. If some day you are sued and you have a conversation with your lawyer about your financial worth as well as how much you are willing to pay to the person suing you - and that information ends up being turned over to the person suing you - you won't be so snide about the importance of attorney client privilege.
> A trendy productivity hack, A.I. note takers are capturing every joke and offhand comment in many meetings. They could also potentially waive attorney-client privilege.
By now everyone knows that AI notes that aren't curated by a human will catch every silly thing that was said in the meeting while omitting the context of the tone or body language. Something as simple as "yeah, right" has vastly different meanings depending on how it was said. In a different context it's already been established that using AI breaks client attorney privilege [0] and this concern has been raised before by law firms [1][2] or the American Bar Association [3] (you can just hit escape before the paywall loads to see the full content). A judge will have to weigh in on this one too.
I don't know what's with the wave of paywalled articles that keep making it to the front page without any workaround included in the submission. Even when you coax the text out of the page source, they're not very insightful to begin with.
If a lawyer takes notes and puts them in a computer, or a cloud drive, or send it over email, they are still covered by attorney-client privilege, right? If they use an AI to do it, it's treated more like a third party no longer covered by the same privilege. If there's no court decision on this it only takes a bit of bad assumption to screw up with using AI.
To be fair, the attorney-client privilege should be completely technology/medium agnostic. If the intention is to have that info stay between client and attorney, nothing should change this.
This is really neat. Is there a way to handle the one app, with one window, being used for multiple tasks?
For example, I have Codex running doing two things at once, and I wish I could have two windows in two spaces (two projects.) Slack has multiple channels.
Both these aren't native macOS apps but I wonder if you can use the macOS tabbing support to at least get this for well-coded native apps?
I don't, but I share concerns. When we see dark patterns, or work on surveillance software (like Facebook), or similar, I think: so many of these devs excuse it as just a job, supporting their family, while actively making the world worse.
So for ethics I'd suggest something that treats software as infrastructure that runs the world, not in an engineering sense but an ethical sense.
I don't yet know how to capture that in three to eight words like your examples.
> a new feature that will boost CPU frequency in bursts when initiating high-priority tasks to speed up app launches and system responsiveness
Great, but what about optimization?
When I click the Start button, the Start menu takes visible time to show and render. I don't think that's a CPU problem. I think that's an implementation problem. There was a perfectly performant Start menu even through Windows 7; what have they done to the taskbar and start menu that makes it so poor now?
and then you click the text box to search (because it doesn't always get focus) and the start menu needs to load another webpage. Sometimes that fails, and the start menu becomes unusable for searching for applications. How the hell does it sometimes take a full minute to open the start menu, type something and press enter, while hoping that the first result you get is the same first result you got when you typed the exact same thing yesterday?
It doesn't exactly answer the question, but I -- someone who doesn't watch video and wants text, articles -- have watched this video twice over. It's worth it :)
reply